Smart court system

Abstract

A Smart-court system, adaptive to constrained sport environment, for enabling real time analysis and debriefing of sport activities is provided herein. The Smart-court system is comprised of: (i) an automatic recording system comprising a plurality of video cameras located in a court, arranged to real-time (RT) recording of a sport session and utilizing automatic calibration and stabilization module; and (ii) a data processing system comprising: (a) a capture module for grabbing a video stream; (b) an objects' detector module arranged to extract during the RT sport session, the objects from the foreground of each frame; (c) an event module for automatically analyzing, the motion and the activities of the tracked objects for automatically identifying and classifying events, creating a synchronized event log and calculating statistics that occurred during the RT sport session; and (d) a presentation module enabling to perform instant debriefing, combined biomechanical and tactical analysis of the video.

What is claimed is: 
     
       1. A computer implemented method of video based sport activity analysis, the method comprising the steps of:
 a) capturing at least one video stream of a sport session using at least one video camera; 
 b) automatically tracking motion of at least one object using the captured video stream; 
 c) based on the tracked motion, automatically identifying occurrence of a sequence of basic events in the captured video stream; 
 d) automatically identifying a complex event that is based on the basic events of the sequence; 
 e) automatically listing at least some of the events in an event log together with data synchronizing between the listed events and the captured video stream; and 
 f) selecting and playing at least one part of the captured video stream to a user, each one of the selected parts showing a respective one of the listed events as captured in the video stream, the part being selected from the video stream using the data synchronizing between the listed events and the captured video stream. 
 
     
     
       2. The method of  claim 1 , further comprising allowing user selection of at least one of the events listed in the event log, wherein said selecting of the at least one part of the captured video stream is further based on the user selection. 
     
     
       3. The method of  claim 1 , further comprising allowing a user to input a query defining a user selection of at least one of the events listed in the event log, wherein said selecting of the at least one part of the captured video stream is further based on the query input by the user. 
     
     
       4. The method of  claim 1 , further comprising allowing user selection of at least one of the events listed in the event log, by event type, wherein said selecting of the at least one part of the captured video stream is further based on the user selection. 
     
     
       5. The method of  claim 1 , further comprising presenting a recommendation to the user upon playing of at least one of the selected parts, the recommendation being selected according to a type of the event presented in the part. 
     
     
       6. The method of  claim 1 , further comprising combining the selected parts into a single video sequence, thereby generating an abridged version of the video stream. 
     
     
       7. The method of  claim 1 , further comprising combining the selected parts into a single video sequence according to a criterion predefined by the user, thereby generating an abridged version of the video stream. 
     
     
       8. The method of  claim 1 , further comprising cutting video sequences in between the selected parts from the captured video stream. 
     
     
       9. The method of  claim 1 , further comprising identifying a mental event based on the tracked motion and on a predefined criterion. 
     
     
       10. The method of  claim 1 , further comprising identifying a side switching event based on the tracked motion and on a predefined criterion. 
     
     
       11. The method of  claim 1 , further comprising identifying the basic events in near real time. 
     
     
       12. The method of  claim 1 , wherein said capturing is carried out using at least two video cameras, each camera simultaneously capturing the sport session in a respective one of the video streams, wherein said tracking is further based on three dimensional (3D) analysis of the simultaneously captured video streams. 
     
     
       13. The method of  claim 1 , further comprising identifying a complex Tennis event based on the tracked motion and on a predefined criterion. 
     
     
       14. The method of  claim 1 , further comprising identifying a Tennis Unforced Error event based on the tracked motion and on a predefined criterion. 
     
     
       15. The method of  claim 1 , further comprising identifying a Tennis Forced Error event based on the tracked motion and on a predefined criterion. 
     
     
       16. The method of  claim 1 , further comprising presenting a list of at least some of the events listed in the event log in sync with said playing of the selected parts on a screen, using the data synchronizing between the listed events and the captured video stream. 
     
     
       17. The method of  claim 1 , further comprising generating an animated video clip based on the tracked motion and playing the animated video clip in sync with said playing of the selected parts, on a screen, using the event log. 
     
     
       18. The method of  claim 1 , further comprising presenting a list of at least some of the events listed in the event log in sync with said playing of the selected parts, on a screen, using the data synchronizing between the listed events and the captured video stream, while allowing the user to jump between two of the played parts by selecting among the events in the presented list. 
     
     
       19. The method of  claim 1 , further comprising generating statistical data based at least on the identified events. 
     
     
       20. The method of  claim 1 , further comprising generating an object profile based at least on the tracked motion. 
     
     
       21. The method of  claim 1 , further comprising generating a player profile based at least on the tracked motion. 
     
     
       22. The method of  claim 1 , further comprising generating a plurality of player profiles based at least on the tracked motion, and matching between players using the generated player profiles. 
     
     
       23. The method of  claim 1 , further comprising generating a player profile based at least on the tracked motion and uploading the generated profile to a social network website. 
     
     
       24. The method of  claim 1 , further comprising automatically identifying a change in a game status based on the tracked movement. 
     
     
       25. The method of  claim 1 , further comprising automatically identifying a start of a game based on the tracked movement. 
     
     
       26. The method of  claim 1 , further comprising automatically identifying an end of a game based on the tracked movement. 
     
     
       27. The method of  claim 1 , further comprising dynamically calculating a game score based on identification of the events.